Pathological characteristics and prognostic indicators of different histopathological types following radical cystectomy: A large cohort with long-term follow-up

INTRODUCTION AND OBJECTIVES: Pathological features and prognostics are better characterized for urothelial cancer (UC) compared to other histological types. Herein we determined differences in pathological features and prognostics among 4 different histological types: UC, UC with metaplasia (UCM), squamous cell carcinoma (SCC) and adenocarcinoma (ADC) utilizing a large cohort of radical cystectomy (RC) patients with long term follow up. METHODS: We retrospectively evaluated 1,280 patients who underwent RC between 1997-2004 at Mansoura, Egypt. Pathological features were compared and prognostics were determined using multivariate analyses; after exclusion of very rare types. RESULTS: We included 1,238 patients (975 males) with median age 58 (29-87) and a median follow up 39 months (0-109). There were significant differences between pathological features and demographics among different subtypes (Table 1, all p <0.05). Schistosmiasis was associated with SCC and ADC (76%) followed by UCM (69%). Nearly all patients with UC and UCM had high-grade disease, versus 41% and 68% patients with SCC and ADC, respectively. Grade was an independent predictor of recurrence only in SCC (HR 1.6, p value = 0.023). Despite the lower incidence of lymphovascular invasion (LVI) in SCC and ADC, it was an independent prognostic in these subtypes (HR 2.1, p < 0.05). Lymph node (LN) involvement was most common in UCM (1/3 of cases) and was the most predictor of recurrence (HR 2.14, p value = 0.012). CONCLUSIONS: Histological subtypes of bladder cancer differ significantly in clinicopathological features and prognostics. SCC and ADC are more associated with younger age, schistosomiasis, and low grade tumor, less LVI and LN involvement. LVI seems to play a greater prognostic role in SCC and ADC. LN metastasis is more common and is associated with worse oncological outcomes in UCM. Future subtype specific prognostic models based on independent predictors can guide selection for multimodal treatments and clinical trials design. (Table Presented).